JP7374201B2 - Machine learning systems and methods, integrated servers, programs, and methods for creating inference models - Google Patents

Description

The present invention relates to a machine learning system and method, an integrated server, an information processing device, a program, and a method for creating an inference model, and particularly relates to a machine learning technique that uses a federated learning mechanism.

When developing medical AI (Artificial Intelligence) using deep learning, it is necessary to train the AI model, but for this learning, training data such as diagnostic images must be transferred from the medical institution to an external development site or development server. I need to bring it out. For this reason, there are currently few medical institutions willing to cooperate in providing learning data. Additionally, even if medical institutions provide training data, there are always privacy risks.

On the other hand, when using the federated learning mechanism proposed in Non-Patent Document 1, learning is performed on the terminal where the data to be learned exists, and a network model that is the learning result on each terminal from a group of these terminals is used. Sends only the weight parameters of the server to the integration server. In other words, in federated learning, only the data of learning results on each terminal is provided from the terminal to the integrated server, without providing learning data to the integrated server.

For this reason, federated learning is a technology that has been attracting attention in recent years because it allows learning without taking the data itself, which requires privacy considerations, outside the organization.

Non-Patent Document 2 reports the results of applying federated learning to the development of medical AI.

H.Brendan, McMahan Eider, Moore Daniel Ramage, Seth Hampson, and Blaise Ag¨ueray Arcas, “Communication-Efficient Learning of Deep Networks from Decentralized Data, arXiv:1602.05629v3 [cs.LG], 28 Feb 2017 Micah J Sheller, G Anthony Reina, Brandon Edwards, Jason Martin, and Spyridon Bakas, “Multi-Institutional Deep Learning Modeling Without Sharing Patient Data: A Feasibility Study on Brain Tumor Segmentation”, arXiv:1810.04304v2 [cs.LG], 22 Oct 2018

When federated learning is used to develop medical AI, there is no need to take out data such as diagnostic images. On the other hand, with the existing federated learning mechanism alone, for example, in situations where an unspecified number of medical institutions participate in learning, it is difficult to achieve the target inference accuracy as early as possible from the start of learning. No specific method has been proposed to achieve this. The content and amount of data held by each medical institution varies, and the learning environment differs from client to client, so the learning outcomes implemented at each client also vary.

With the existing federated learning mechanism alone, when an unspecified number of medical institutions participate in learning, which client's learning results are optimally used to create a master model that is an integrated model? There is no such indicator.

Therefore, if a combination of clients is selected at random from among a large number of clients, the inference accuracy of the integrated model will not reach the target if the learning environments of the clients that make up the selected client group are biased. Alternatively, it may take a large amount of learning time to reach the target accuracy.

The present invention was made in view of these circumstances, and implements a federated learning mechanism that allows AI models to learn without taking personal information such as diagnostic images that require privacy considerations from the medical institution side. The present invention aims to provide a machine learning system and method, an integrated server, an information processing device, a program, and a method for creating an inference model that can quickly improve the inference accuracy of a model.

A machine learning system according to one aspect of the present disclosure is a machine learning system including a plurality of client terminals and an integrated server, each of the plurality of client terminals being stored in a data storage device of a medical institution. The integrated server includes a learning processing unit that executes machine learning of the learning model using data as learning data, and a transmitting unit that sends the learning results of the learning model to the integrated server. a synchronization processing unit that synchronizes the learning model of each client terminal with the master model before the client terminals of the client terminals learn the respective learning models; a receiving unit that receives the learning results from the plurality of client terminals; A process of searching for a combination of client terminals for which the inference accuracy of a master model candidate that is created by integrating the learning results of combinations of client terminals that are part of the client terminals of The present invention is a machine learning system including a client combination optimization processing unit that performs at least one of the processes of searching for a maximizing combination of client terminals.

According to this aspect, a combination of client terminals is extracted from among a plurality of client terminals participating in learning, and a master model candidate is created by integrating the learning results of a group of client terminals belonging to the combination. Then, processing is performed to verify the inference accuracy of the created master model candidate and search for the optimal combination of client terminals that is effective in improving accuracy. The term "optimal" herein is not limited to the most suitable one, but includes what is understood as one of the most suitable. In other words, the optimal combination includes the concepts of both the optimal solution of the combination and an approximate solution close to the optimal solution.

A combination of client terminals in which the inference accuracy of the master model candidate satisfies the target accuracy is understood to be one of the optimal combinations. The statement "meeting the target accuracy" includes achieving the target accuracy. For example, achieving an inference accuracy that exceeds an accuracy target value indicating a target accuracy is one aspect of "meeting the target accuracy." If a master model candidate is found that satisfies the target accuracy set as the goal of inference accuracy, the combination of client terminals used to create the master model candidate may be regarded as one of the optimal combinations.

Further, the combination of client terminals that maximizes the inference accuracy of the master model candidate is understood to be one of the optimal combinations. The description "a combination of client terminals that maximizes inference accuracy" is not limited to a combination that maximizes inference accuracy, but includes a combination that is understood to be one of the combinations that provides inference accuracy that is close to the maximum. The combination of client terminals used to create the master model candidate with the highest inference accuracy among the multiple master model candidates created in the search process is understood to be one of the "combinations of client terminals that maximize inference accuracy." It can be considered as one of the optimal combinations.

According to this aspect, it is possible to extract a combination of client terminals that provides better learning accuracy from among a plurality of client terminals. This makes it possible to create a model with higher inference accuracy than the initial master model at a relatively early stage after the start of learning.

The "multiple client terminals" may be an unspecified number of client terminals. The client terminal may include a "data storage device of a medical institution", or the "data storage device of a medical institution" and the "client terminal" may be separate devices.

The client combination optimization processing unit performs a first search process to search for a combination of client terminals for which the inference accuracy of the master model candidate satisfies a target accuracy, and a search process for a combination of client terminals that maximizes the inference accuracy of the master model candidate. The configuration may be such that only one of the second search processes is executed, or the configuration may be configured to execute both of the search processes. For example, the client combination optimization processing unit may be configured to perform the second search process when the first search process fails to find a combination of client terminals that satisfies the target accuracy. .

In the machine learning system according to another aspect of the present disclosure, the client combination optimization processing unit integrates learning results of the client clusters with a client cluster creation unit that creates a client cluster that is a combination of client terminals from a plurality of client terminals. a master model candidate creation unit that creates a master model candidate, and an accuracy evaluation unit that detects a master model candidate whose inference accuracy exceeds an accuracy target value by evaluating the inference accuracy of the master model candidate. can do.

In the machine learning system according to still another aspect of the present disclosure, the accuracy evaluation unit compares the inference result output from the master model candidate by inputting the verification data into the master model candidate and the correct data of the verification data. The configuration may include an inference accuracy calculation unit that calculates the inference accuracy of the master model candidate, and an accuracy target value comparison unit that compares the inference accuracy of the master model candidate with an accuracy target value.

In the machine learning system according to still another aspect of the present disclosure, the accuracy evaluation unit is configured to perform evaluation based on a comparison between an instantaneous value of the inference accuracy of the master model candidate and an accuracy target value, or in each learning iteration of the master model candidate. It can be configured to determine whether the inference accuracy of the master model candidate exceeds the accuracy target value based on a comparison between the statistical value of the inference accuracy and the accuracy target value.

The "statistical value" is a statistical value calculated using a statistical algorithm, and may be a representative value such as an average value or a median value.

In a machine learning system according to still another aspect of the present disclosure, the client combination optimization processing unit performs a process of extracting a specified number of client terminals from among the plurality of client terminals and creating a combination of client terminals. , based on the process of integrating the learning results for each combination of client terminals to create a master model candidate for each combination, and the comparison result of the inference accuracy of each master model candidate created for each combination with the accuracy target value. , and a process of searching for a combination of client terminals that exceeds the accuracy target value.

In a machine learning system according to still another aspect of the present disclosure, each of the plurality of client terminals may be a terminal installed within a medical institution network of a different medical institution.

In a machine learning system according to still another aspect of the present disclosure, the integrated server may be configured to be installed within a medical institution network or outside a medical institution network.

In a machine learning system according to still another aspect of the present disclosure, the learning results sent from the client terminal to the integrated server may include weight parameters of the learning model after learning.

In the machine learning system according to still another aspect of the present disclosure, the data used as learning data includes at least one type of data among two-dimensional images, three-dimensional images, moving images, time series data, and document data. It can be configured as follows.

In a machine learning system according to still another aspect of the present disclosure, each of the learning model, master model, and master model candidate may be configured using a neural network.

An appropriate network model is applied depending on the type of training data and data input during inference.

In a machine learning system according to still another aspect of the present disclosure, data used as learning data includes a two-dimensional image, a three-dimensional image, or a moving image, and each model of the learning model, master model, and master model candidate is It may be constructed using a convolutional neural network.

In a machine learning system according to still another aspect of the present disclosure, data used as learning data includes time series data or document data, and each model of the learning model, master model, and master model candidate is a recurrent neural network. It may be configured using

In the machine learning system according to still another aspect of the present disclosure, the integrated server determines the inference accuracy of the master model candidate created for each combination of client terminals and the combination of client terminals used by the master model candidate. The configuration may further include an information storage unit that stores information indicating a correspondence relationship, such as whether the information has been created or not.

In a machine learning system according to still another aspect of the present disclosure, the integrated server further includes a display device that displays inference accuracy in each learning iteration of each master model candidate created for each combination of client terminals. can do.

A machine learning system according to still another aspect of the present disclosure may further include a verification data storage unit that stores verification data used when evaluating inference accuracy of a master model candidate.

The verification data storage unit may be included in the integrated server, or may be an external storage device connected to the integrated server.

A machine learning method according to another aspect of the present disclosure is a machine learning method using a plurality of client terminals and an integrated server. By synchronizing the side learning model with the trained master model stored in the integrated server, and by allowing each of the multiple client terminals to synchronize the data stored in the respective data storage devices of different medical institutions. Execute machine learning of the learning model using the training data, each of the multiple client terminals send the learning results of the learning model to the integrated server, and the integrated server performs the learning of each learning model from the multiple client terminals. a process of receiving the results and searching for a combination of client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of combinations of client terminals that are part of the plurality of client terminals satisfies a target accuracy; and performing at least one of a process of searching for a combination of client terminals that maximizes the inference accuracy of the master model candidate.

An integrated server according to another aspect of the present disclosure is an integrated server that is connected to a plurality of client terminals via a communication line, and includes a master model storage unit that stores a learned master model, and a master model storage unit that stores a learned master model, and A synchronization processing unit that synchronizes the learning model of each client terminal with the master model before the terminal learns each learning model, a receiving unit that receives the learning results from multiple client terminals, and a synchronization processing unit that synchronizes the learning model of each client terminal with the master model, The process of searching for a combination of client terminals whose inference accuracy of a master model candidate created by integrating the learning results of combinations of client terminals that are part of the terminal satisfies the target accuracy, and maximizing the inference accuracy of the master model candidate. The integrated server includes a client combination optimization processing unit that performs at least one of the processes of searching for a combination of client terminals.

An integrated server according to another aspect of the present disclosure is an integrated server connected to a plurality of client terminals via a communication line, and includes a first processor and a first program executed by the first processor. a first computer-readable medium that is a recorded non-transitory tangible object; the first processor stores the trained master model on the first computer-readable medium according to instructions of the first program; Before having multiple client terminals learn each learning model, synchronize the learning model on each client terminal with the master model, and receive the learning results from multiple client terminals. A process of searching for a combination of client terminals whose inference accuracy of a master model candidate, which is created by integrating learning results of combinations of client terminals that are part of multiple client terminals, satisfies a target accuracy; This is an integrated server that executes processing including at least one of the processing of searching for a combination of client terminals that maximizes inference accuracy.

In the integrated server according to still another aspect of the present disclosure, the first processor extracts some of the client terminals from the plurality of client terminals and creates a client cluster, which is a combination of the client terminals, according to the instructions of the first program. The configuration can be configured to execute processing including creating a master model candidate, and creating a master model candidate by integrating learning results of client clusters.

An information processing apparatus according to another aspect of the present disclosure is an information processing apparatus used as one of a plurality of client terminals connected to the integrated server according to one aspect of the present disclosure via a communication line, The learning model synchronized with the master model stored in is used as the learning model in the initial state before learning starts, and the machine learning of the learning model is performed using the data stored in the data storage device of the medical institution as the learning data. The information processing apparatus includes a learning processing unit that executes the learning process, and a transmitting unit that transmits the learning results of the learning model to the integrated server.

An information processing apparatus according to another aspect of the present disclosure is an information processing apparatus used as one of a plurality of client terminals connected to the integrated server according to one aspect of the present disclosure via a communication line, a second computer-readable medium that is a non-transitory tangible object having recorded thereon a second program executed by the second processor, the second processor executing instructions of the second program; Accordingly, the learning model synchronized with the master model stored in the integrated server is used as the learning model in the initial state before learning starts, and the learning model is created using data stored in the data storage device of the medical institution as the learning data. An information processing device that executes processing including executing machine learning of a learning model and transmitting learning results of a learning model to an integrated server.

A program according to another aspect of the present disclosure is a program for causing a computer to function as one of a plurality of client terminals connected to the integrated server according to one aspect of the present disclosure via a communication line, The learning model synchronized with the master model stored in is used as the learning model in the initial state before learning starts, and the machine learning of the learning model is performed using the data stored in the data storage device of the medical institution as the learning data. This is a program that allows a computer to implement the functions to execute and the function to send the learning results of the learning model to the integrated server.

A program according to another aspect of the present disclosure is a program for causing a computer to function as an integrated server connected to a plurality of client terminals via a communication line, the program storing a learned master model in the computer. A function to synchronize the learning model on each client terminal with the master model before having multiple client terminals learn each learning model, and a function to receive each learning result from multiple client terminals. , a process of searching for a combination of client terminals for which the inference accuracy of a master model candidate created by integrating the learning results of combinations of client terminals that are part of multiple client terminals satisfies a target accuracy, and inference of the master model candidate. This is a program for realizing a function of performing at least one of the processes of searching for a combination of client terminals that maximizes accuracy.

A method for creating an inference model according to another aspect of the present disclosure is a method for creating an inference model by performing machine learning using a plurality of client terminals and an integrated server. Before training a learning model on each client terminal, it is necessary to synchronize the learning model on each client terminal side with the trained master model stored on the integrated server, and to synchronize the learning model on each client terminal side with the trained master model stored on the integrated server. Executing machine learning of the learning model using data stored in each data storage device as learning data, and each of the multiple client terminals transmitting the learning results of the learning model to the integrated server. , the goal is for the integrated server to receive learning results from multiple client terminals, and for the inference accuracy of the master model candidate created by integrating the learning results of combinations of client terminals that are part of the multiple client terminals. and the process of searching for a combination of client terminals that maximizes the inference accuracy of the master model candidate, and the inference accuracy that satisfies the target accuracy. A master model candidate that has achieved this, or a master model candidate created using the combination of client terminals used to create the model with the highest inference accuracy among multiple master model candidates created through the search process. This is a method for creating an inference model, which includes the steps of: creating an inference model with higher inference accuracy than the master model based on the master model;

The method for creating an inference model is understood as an invention of a method for manufacturing an inference model. The term "inference" includes the concepts of prediction, estimation, classification, and discrimination. The inference model may also be referred to as an "AI model."

According to the present invention, an optimal combination of client terminals to be used for integrating learning results can be found from among a plurality of client terminals. This makes it possible to perform learning efficiently and improve the inference accuracy of the model at an early stage.

FIG. 1 is a conceptual diagram showing an overview of a machine learning system according to an embodiment of the present invention. FIG. 2 is a diagram schematically showing a system configuration example of a machine learning system according to an embodiment of the present invention. FIG. 3 is a block diagram showing an example of the configuration of the integrated server. FIG. 4 is a block diagram showing a configuration example of a CAD (Computer Aided Detection/Diagnosis) server, which is an example of a client. FIG. 5 is a flowchart showing an example of the operation of the client terminal based on the local learning management program. FIG. 6 is a flowchart showing an example of the operation of the integrated server 30 based on the learning client combination optimization program 33. FIG. 7 is a flowchart illustrating an example of a process for evaluating the inference accuracy of a master model candidate in the integrated server. FIG. 8 is a block diagram showing an example of the hardware configuration of a computer.

Preferred embodiments of the present invention will be described below with reference to the accompanying drawings.

《Overview of machine learning system》
FIG. 1 is a conceptual diagram showing an overview of a machine learning system according to an embodiment of the present invention. The machine learning system 10 is a computer system that performs machine learning using a federated learning mechanism. Machine learning system 10 includes multiple clients 20 and an integrated server 30. Federated learning is sometimes called "federated learning,""cooperativelearning," or "federated learning."

Each of the plurality of clients 20 shown in FIG. 1 represents a medical institution terminal installed on a network within a medical institution such as a hospital. Here, a "terminal" refers to a computing resource that exists within a network that can safely access data within a medical institution, and the terminal does not need to be physically located within the medical institution. The client 20 is an example of a "client terminal" in the present disclosure. A computer network within a medical institution is called a "medical institution network."

It is assumed that each client 20 exists for each data group to be trained by the AI model. Here, "each data group" may be understood as "each medical institution" in which the data group used for learning the AI model is held. That is, it is assumed that approximately one client exists for one medical institution.

In order to distinguish and display each of the plurality of clients 20, notations such as "Client 1" and "Client 2" are used in FIG. 1 and subsequent drawings. The number appended to “Client” is an index as an identification number that identifies each client 20. In this specification, the client 20 whose index is m is referred to as "client CLm." For example, client CL1 represents "Client 1" in FIG. m corresponds to a client ID number (identification number). When the total number of clients 20 managed by the integrated server 30 is M, m represents an integer from 1 to M, inclusive. In FIG. 1, clients 20 from m=1 to N+1 are illustrated. N represents an integer of 2 or more. The entire set of M clients 20 participating in learning is referred to as a "learning client group" or a "client population."

Each client 20 has local data LD in a storage device local to the client. The local data LD is a data group accumulated at the medical institution to which the client 20 belongs.

Each client 20 includes a local learning management program that is a distributed learning client program. Each client 20 runs an iteration of learning the local model LM using local data LD local to the client according to the local learning management program.

The local model LM is, for example, an AI model for medical image diagnosis that is incorporated into a CAD system. The term "CAD" includes the concepts of both Computer Aided Detection (CADe) and Computer Aided Diagnosis (CADx). The local model LM is configured using, for example, a hierarchical multilayer neural network. In the local model LM, the weight parameters of the network are updated by deep learning using the local data LD as learning data. The weight parameters include filter coefficients (weights of connections between nodes) of filters used for processing each layer and biases of nodes. The local model LM is an example of a "client terminal side learning model" in the present disclosure.

Note that a "neural network" is a mathematical model for information processing that simulates the structure of the brain's nervous system. Processing using a neural network can be realized using a computer. A processing unit including a neural network may be configured as a program module.

As the network structure of the neural network used for learning, an appropriate network structure is adopted depending on the type of data used for input. AI models for medical image diagnosis can be configured using, for example, various convolutional neural networks (CNNs) having convolutional layers. AI models that handle time series data, document data, etc. can be configured using, for example, various types of recurrent neural networks (RNNs).

The plurality of clients 20 are connected to the integrated server 30 via a communication network. The integration server 30 performs a process of acquiring learning results from a plurality of clients 20, a process of integrating learning results for combinations of clients 20 extracted from a population to create a master model candidate MMC, and a process of creating a master model candidate MMC. A process of evaluating the inference accuracy of the MMC and a process of optimizing the combination of clients 20 based on the evaluation result of the inference accuracy are performed. In this specification, a client group that is a combination of clients 20 used to create a master model candidate MMC is referred to as a "client cluster."

The location of the integrated server 30 only needs to be on a computer network to which the entity developing the AI model has access rights, and the form of the server does not matter, such as a physical server or a virtual server. The integrated server 30 may be installed within the medical institution network or may be installed outside the medical institution network. For example, the integrated server 30 may be installed within a company that develops medical AI that is geographically remote from the medical institution or on the cloud.

In FIG. 1, clients CL1, CL2 and CL3 belong to the same client cluster.

The arrow extending from the left side of the circle surrounding the display "Federated Avg" in FIG. 1 indicates that data of the learned local model LM is being transmitted from each client 20 belonging to the same client cluster. The data of the local model LM as a learning result provided from each client 20 to the integrated server 30 may be a weight parameter of the local model LM after learning.

The circle surrounding the display "Federated Avg" represents the process of integrating learning results. In this process, the weights sent from each client 20 are integrated by averaging or the like, and a master model candidate MMC, which is an integrated model, is created. The integration processing method is not limited to simple arithmetic averaging, but also takes into account factors such as the attributes of the client 20, past integration results, the number of data for each medical institution used for relearning, and the level of medical institutions evaluated by humans. Based on this, it may be weighted and integrated.

In FIG. 1, "Master model" shown at the tip of the arrow extending to the right of the circle surrounding the display "Federated Avg" indicates a master model candidate MMC created from the client cluster.

The integrated server 30 evaluates the inference accuracy of the master model candidate MMC using verification data prepared in advance. The verification data may be stored in an internal storage device of the integrated server 30, or may be stored in an external storage device connected to the integrated server 30.

The integrated server 30 includes a learning client combination optimization program 33 and a database 36.

The learning client combination optimization program 33 stores the inference accuracy of the master model candidate MMC in a data storage unit such as the database 36 . The data storage unit may be a storage area of a storage device within the integrated server 30, or may be a storage area of an external storage device connected to the integrated server 30. The training client combination optimization program 33 also stores information on which combination of clients 20 (client group) was used to create the master model candidate MMC for which the inference accuracy has been calculated, in a data storage unit such as the database 36. save.

The learning client combination optimization program 33 compares the inference accuracy of the master model candidate MMC with the accuracy target value, and determines whether a master model candidate MMC with inference accuracy exceeding the accuracy target value is found or the number of iterations has reached the upper limit iteration. The creation of a master model candidate MMC and the evaluation of its inference accuracy are repeated by changing the combination of clients 20 until the number reaches the number, and the combination of learning results of the clients 20 that improves the inference accuracy of the master model candidate MMC is created. (i.e. client combinations). Note that among the client population, there may be clients 20 that are not used for creating the master model candidate MMC.

If the integration server 30 does not find a client combination for which the inference accuracy of the master model candidate MMC exceeds the accuracy target value within the time limit, the integrated server 30 further advances the learning of the client 20.

《Overview of machine learning methods》
An example of a machine learning method using the machine learning system 10 according to an embodiment of the present invention will be described. The machine learning system 10 operates according to steps 1 to 11 shown below.

[Step 1] As shown in Figure 1, data for federated learning is created on the medical institution terminal (client 20) in the medical institution's computer network where the data group that the AI model should learn exists. The distributed learning client program is running.

[Step 2] Before each of the multiple clients 20 starts learning, the integrated server 30 synchronizes the latest version of the master model used for learning with the local model LM on each client 20. The master model is a trained AI model.

[Step 3] After synchronizing with the latest version of the master model, each client 20 performs learning on its own terminal using the local data LD existing within the medical institution, and performs the learning process for the specified number of iterations. turn. The local data LD used as learning data may be, for example, a medical image and information accompanying this. The "accompanying information" may include information equivalent to the teacher signal. The number of iterations may be a fixed value, but more preferably, the learning iterations are repeated until the inference accuracy has improved by a specified percentage or more.

[Step 4] After each client 20 completes learning, it transmits the learning results to the integrated server 30. The learning results sent from the client 20 to the integrated server 30 may be weight parameters of the local model LM after learning. Note that the data of the weight parameters after learning transmitted from the client 20 to the integrated server 30 may be a difference from the weight parameters of the latest version of the master model synchronized with the integrated server 30.

A document attached to a medical device or the like that is a client 20 that uses the functions according to the present embodiment states that learning is performed as background processing within a range that does not interfere with medical work. Additionally, the attached document states that the learning data used is internal to the medical institution, and that the only data sent externally is the weight parameters after learning, and that no data that could identify individuals will be sent. Ru.

[Step 5] The learning client combination optimization program 33 running on the integrated server 30 extracts the specified number W of clients 20 from the client population, and uses the learning results received from these clients 20. They are integrated to create a master model candidate MMC. The learning client combination optimization program 33 stores client combination information indicating from which combination of clients 20 the created master model candidate MMC is created in a data storage unit such as the database 36 .

[Step 6] The learning client combination optimization program 33 verifies the inference accuracy of the created master model candidate MMC. Accuracy verification is performed on verification data. That is, the learning client combination optimization program 33 causes the master model candidate MMC to perform inference using the verification data existing in the integrated server 30 as input, and compares the inference result with the correct data. The inference accuracy is calculated, and the inference accuracy of the master model candidate MMC is stored in a data storage unit such as the database 36. The database 36 is an example of an "information storage unit" in the present disclosure.

[Step 7] The training client combination optimization program 33 changes the combinations of clients 20 until the master model candidate's inference accuracy exceeds the accuracy target value or until the upper limit number of iterations is reached. The process of creating model candidates and calculating their inference accuracy is repeated.

When performing this combination search, the combination of clients 20 that maximizes the inference accuracy of the master model candidate may be searched for by brute force changing the combinations from the client population, but it is more preferable to use the weight parameters of the local model LM. An optimization method suitable for updating the weight parameters of the target model is used for the combinatorial search, such as the same optimization method used to update the weight parameters of the target model.

This search problem is a problem of searching for a combination of clients 20 used to create a master model candidate MMC, that is, changing the weight parameters of the network of the target master model candidate MMC and searching for a direction that maximizes inference accuracy. This can be said to be a problem. This is a problem similar to the optimization problem of updating weight parameters during model learning, so it is more preferable to use the same stochastic gradient descent method as when learning the local model LM, or use other weight parameters of the target model. It is preferable to adopt an optimization method suitable for updating.

[Step 8] If a combination of clients 20 for which the inference accuracy of the master model candidate MMC exceeds the accuracy target value is found, the learning process may be terminated at that stage.

[Step 9] In Step 7, if a combination of clients 20 for which the inference accuracy of the master model candidate MMC exceeds the accuracy target value within the upper limit number of iterations is not found, then The master model candidate MMC with the best inference accuracy is synchronized with the client group, and steps 2 to 9 are repeated.

This makes it possible to find an optimal combination of clients 20 that maximizes the inference accuracy of the master model candidate at an early stage, and to create an inference model with inference accuracy that exceeds the target accuracy value. The machine learning method using the machine learning system 10 according to this embodiment is understood as a method for creating an inference model.

《System configuration example》
Next, an example of a specific configuration of the machine learning system 10 will be described. FIG. 2 is a diagram schematically showing a system configuration example of the machine learning system 10 according to the embodiment of the present invention. First, an example of the medical institution network 50 will be explained. In order to simplify the illustration, FIG. 2 shows an example in which a medical institution network 50 with the same system configuration is installed in each of a plurality of medical institutions, but a medical institution network with a different system configuration for each medical institution is shown. may be constructed.

The medical institution network 50 includes a CT (Computed Tomography) device 52, an MRI (Magnetic Resonance Imaging) device 54, a CR (Computed Radiography) device 56, a PACS (Picture Archiving and Communication Systems) server 58, and a CAD server 60. , a terminal 62, and a local communication line 64.

Note that the medical institution network 50 is not limited to the CT device 52, MRI device 54, and CR device 56 illustrated in FIG. Even if it includes at least one or a combination of imaging devices, angiography X-ray diagnostic devices, ultrasound diagnostic devices, PET (Positron Emission Tomography) devices, endoscope devices, mammography devices, and various other examination devices (modalities). good. There may be various combinations of types of testing devices connected to the medical institution network 50 for each medical institution.

The PACS server 58 is a computer that stores and manages various data, and includes a large-capacity external storage device and database management software. The PACS server 58 communicates with other devices via the local communication line 64 and sends and receives various data including image data. The PACS server 58 receives image data and other various data generated by each inspection device such as the CT device 52, MRI device 54, and CR device 56 via the in-house communication line 64, and stores it in a large-capacity external storage device or the like. Save and manage on a recording medium.

Note that the storage format of image data and the communication between each device via the local communication line 64 are based on a protocol such as DICOM (Digital Imaging and Communication in Medicine). PACS server 58 may be a DICOM server that operates according to DICOM specifications. Data stored in the PACS server 58 can be used as learning data. Further, learning data created based on data stored in the PACS server 58 can also be stored in the CAD server 60. The PACS server 58 is an example of a "data storage device of a medical institution" in the present disclosure. Further, the CAD server 60 may function as a "data storage device of a medical institution" in the present disclosure.

The CAD server 60 corresponds to the client 20 described in FIG. The CAD server 60 has a communication function for communicating with the integrated server 30 and is connected to the integrated server 30 via a wide area communication line 70. The CAD server 60 can acquire data from the PACS server 58 and the like via the local communication line 64. The CAD server 60 includes a local learning management program for executing learning of the local model LM on the CAD server 60 using the data group stored in the PACS server 58. The CAD server 60 is an example of a "client terminal" in the present disclosure.

Various data stored in the database of the PACS server 58 and various information including the inference results by the CAD server 60 can be displayed on the terminal 62 connected to the local communication line 64.

The terminal 62 may be a display terminal called a PACS viewer or a DICOM viewer. A plurality of terminals 62 may be connected to the medical institution network 50. The form of the terminal 62 is not particularly limited, and may be a personal computer, a workstation, a tablet terminal, or the like.

As shown in FIG. 2, a medical institution network having a similar system configuration is constructed for each of a plurality of medical institutions. The integrated server 30 communicates with a plurality of CAD servers 60 via a wide area communication line 70. The wide area communication line 70 is an example of a "communication line" in the present disclosure.

<<Configuration example of integrated server 30>>
FIG. 3 is a block diagram showing an example of the configuration of the integrated server 30. The integrated server 30 can be realized by a computer system configured using one or more computers. The integrated server 30 is realized by installing and executing a program on a computer.

Integrated server 30 includes a processor 302 , a non-transitory tangible computer readable medium 304 , a communication interface 306 , an input/output interface 308 , a bus 310 , an input device 314 , and a display device 316 . Processor 302 is an example of a "first processor" in this disclosure. Computer-readable medium 304 is an example of a "first computer-readable medium" in this disclosure.

Processor 302 includes a CPU (Central Processing Unit). The processor 302 may include a GPU (Graphics Processing Unit). Processor 302 is connected to computer readable media 304, communication interface 306, and input/output interface 308 via bus 310. Input device 314 and display device 316 are connected to bus 310 via input/output interface 308 .

Computer-readable medium 304 includes memory, which is a main storage device, and storage, which is an auxiliary storage device. Computer-readable medium 304 may be, for example, a semiconductor memory, a hard disk drive (HDD) device, a solid state drive (SSD) device, or a combination of these.

The integrated server 30 is connected to a wide area communication line 70 (see FIG. 2) via a communication interface 306.

Computer-readable medium 304 includes a master model storage section 320, a verification data storage section 322, and a database 36. The master model storage unit 320 stores data of the latest version of the master model MM. The verification data storage unit 322 stores a plurality of verification data TD used when verifying the inference accuracy of the integrated model created by the master model candidate creation unit 334. The verification data TD is a combination of input data and correct answer data, and is also called test data. The verification data TD may be, for example, data provided by a university or the like.

The computer readable medium 304 stores various programs and data including a synchronization program 324 and a learning client combination optimization program 33. The synchronization program 324 is a program for providing data of the master model MM to each client 20 via the communication interface 306 and synchronizing each local model LM with the master model MM. When the processor 302 executes the instructions of the synchronization program 324, the computer functions as a synchronization processing section. Note that the synchronization program 324 may be incorporated as a program module of the learning client combination optimization program 33.

When the processor 302 executes the instructions of the learning client combination optimization program 33, the computer functions as a client combination optimization processing section 330. The client combination optimization processing section 330 includes a client cluster extraction section 332, a master model candidate creation section 334, and an inference accuracy evaluation section 340. The inference accuracy evaluation unit 340 includes an inference unit 342, an inference accuracy calculation unit 344, and an accuracy target value comparison unit 346.

The client cluster extraction unit 332 extracts a combination of clients 20 to be used for creating a master model candidate MMC from among a plurality of clients 20, and creates a client cluster. For example, the client cluster extraction unit 332 extracts a specified number of clients 20 from the client population randomly or according to a predetermined algorithm to create a client cluster. The number of clients in a client cluster may be a fixed value specified by a program, or may be one of the variables when optimizing the client combination.

The client cluster extraction unit 332 can create a plurality of client clusters with different combinations of clients 20 from the client population. A plurality of client clusters that are a combination of various clients 20 created by the client cluster extraction unit 332 may include some of the clients 20 that make up each client cluster. When the client cluster extraction unit 332 creates multiple client clusters, it is not necessary to allocate all clients in the client population to any one of the client clusters, and the learning results of some clients 20 are not used in the integration process. You can.

Creation of client clusters by the client cluster extraction unit 332 may be performed before each client 20 starts learning, or after the start of learning. This may be done after receiving the. Communication interface 306 is an example of a "receiving unit" in the present disclosure. The client cluster extraction unit 332 is an example of a “client cluster creation unit” in the present disclosure.

The client cluster extraction unit 332 stores in the database 36 information indicating the correspondence between the information of the clients 20 belonging to each client cluster and the master model candidate MMC created for each client cluster.

The master model candidate creation unit 334 integrates the learning results for each client cluster and creates a master model candidate MMC. Information indicating the correspondence relationship regarding which client cluster the master model candidate MMC was created based on is stored in the database 36.

The inference accuracy evaluation unit 340 verifies and evaluates the inference accuracy of the master model candidate MMC created for each client cluster.

The inference unit 342 inputs the verification data TD to the master model candidate MMC and executes inference using the master model candidate MMC. The inference accuracy calculation unit 344 compares the inference result of the master model candidate MMC obtained from the inference unit 342 with the correct data, and calculates the inference accuracy of the master model candidate MMC. For example, the correct data includes data in which the number of lesions and correct clinical findings are added together with image data. The inference accuracy calculation unit 344 performs accuracy verification multiple times through comparison with verification data. The inference accuracy calculation unit 344 may calculate an accuracy average value of the master model candidate from the results of multiple accuracy verifications, and evaluate this accuracy average value as the inference accuracy of the master model candidate. The inference accuracy calculated by the inference accuracy calculation unit 344 is stored in the database 36.

The accuracy target value comparison unit 346 selects the inference accuracy of the model with the highest inference accuracy among the plurality of created master model candidates, compares the inference accuracy with the target accuracy (accuracy target value), It is determined whether a master model candidate with inference accuracy exceeding the accuracy target value has been obtained. The accuracy target value is set, for example, to a higher accuracy than the inference accuracy of the latest version of the master model MM, and is set to a level of accuracy that can be commercialized instead of the master model MM.

The synchronization program 324 and the learning client combination optimization program 33 are examples of the "first program" in the present disclosure.

Furthermore, the computer functions as a display control unit 350 by the processor 302 executing instructions of the display control program. The display control unit 350 generates display signals necessary for display output to the display device 316 and controls the display of the display device 316.

The display device 316 is configured by, for example, a liquid crystal display, an organic electro-luminescence (OEL) display, a projector, or an appropriate combination thereof. The input device 314 is configured by, for example, a keyboard, a mouse, a touch panel, or other pointing device, a voice input device, or an appropriate combination thereof. Input device 314 accepts various inputs from an operator. Note that the display device 316 and the input device 314 may be configured integrally by using a touch panel.

The display device 316 can display the inference accuracy in each learning iteration of each of the plurality of master model candidates MMC.

<<Configuration example of CAD server 60>>
FIG. 4 is a block diagram showing a configuration example of a CAD server 60, which is an example of the client 20. The CAD server 60 can be realized by a computer system configured using one or more computers. The CAD server 60 is realized by installing and executing a program on a computer.

CAD server 60 includes a processor 602, a non-transitory tangible computer readable medium 604, a communication interface 606, an input/output interface 608, a bus 610, an input device 614, and a display device 616. The hardware configuration of the CAD server 60 may be similar to the hardware configuration of the integrated server 30 described in FIG. 3. That is, the respective hardware configurations of processor 602, computer readable medium 604, communication interface 606, input/output interface 608, bus 610, input device 614, and display device 616 in FIG. , communication interface 306, input/output interface 308, bus 310, input device 314, and display device 316.

The CAD server 60 is an example of an "information processing device" in the present disclosure. Processor 602 is an example of a "second processor" in this disclosure. Computer-readable medium 604 is an example of a "second computer-readable medium" in this disclosure.

The CAD server 60 is connected to the learning data storage section 80 via a communication interface 606 or an input/output interface 608. The learning data storage unit 80 is configured to include a storage that stores learning data used by the CAD server 60 to perform machine learning. "Learning data" is training data used for machine learning, and is synonymous with "learning data" or "training data." The learning data stored in the learning data storage unit 80 is the local data LD described in FIG. 1. The learning data storage unit 80 may be the PACS server 58 described in FIG. 2. The learning data storage unit 80 is an example of a "data storage device of a medical institution" in the present disclosure.

Here, an example will be described in which the learning data storage unit 80 and the CAD server 60 that executes learning processing are configured as separate devices, but these functions may be realized by a single computer. The processing function may be shared between two or more computers.
The computer readable medium 604 of the CAD server 60 shown in FIG. 4 stores various programs and data including a local learning management program 630 and a diagnostic support program 640. When the processor 602 executes the instructions of the local learning management program 630, the computer executes the synchronization processing unit 631, learning data acquisition unit 632, local model LM, error calculation unit 634, optimizer 635, learning result storage unit 636, and transmission processing. 637. The local learning management program 630 is an example of a "second program" in the present disclosure.

The synchronization processing unit 631 communicates with the integrated server 30 via the communication interface 606 and synchronizes the master model MM in the integrated server 30 and the local model LM in the CAD server 60.

The learning data acquisition unit 632 acquires learning data from the learning data storage unit 80. The learning data acquisition section 632 may be configured to include a data input terminal that takes in data from the outside or from another signal processing section within the device. The learning data acquisition unit 632 also includes a communication interface 306, an input/output interface 308, a media interface that reads and writes from a portable external storage medium such as a memory card (not shown), or an appropriate combination of these aspects. may be configured.

The learning data acquired via the learning data acquisition unit 632 is input to the local model LM as a learning model.

The error calculation unit 634 calculates the error between the predicted value indicated by the score output from the local model LM and the correct data. The error calculation unit 634 evaluates the error using a loss function. The loss function may be, for example, cross entropy or mean squared error.

The optimizer 635 performs a process of updating the weight parameters of the local model LM from the calculation results of the error calculation unit 634. The optimizer 635 uses the error calculation result obtained from the error calculation unit 634 to perform calculation processing to determine the update amount of the weight parameter of the local model LM, and according to the calculated update amount of the weight parameter, the weight parameter of the local model LM. Update processing and. The optimizer 635 updates weight parameters based on an algorithm such as error backpropagation.

The CAD server 60 incorporating the local learning management program 630 functions as a local learning device that executes machine learning on the CAD server 60 using the local data LD as learning data. The CAD server 60 reads learning data, which is local data LD, from the learning data storage unit 80 and executes machine learning. The CAD server 60 can read learning data and update weight parameters in units of mini-batches that are a collection of a plurality of learning data. A processing unit including the learning data acquisition unit 632, local model LM, error calculation unit 634, and optimizer 635 is an example of a “learning processing unit” in the present disclosure.

The local learning management program 630 repeats the learning process until the learning end condition is met. After the learning termination condition is satisfied, the weight parameters of the local model LM as the learning results are stored in the learning result storage unit 636.

The transmission processing unit 637 performs a process of transmitting the learning results to the integrated server 30. The weight parameters of the local model LM after learning stored in the learning result storage unit 636 are sent to the integrated server 30 via the communication interface 606 and the wide area communication line 70 (see FIG. 2). The transmission processing unit 637 and the communication interface 606 are an example of a “transmission unit” in the present disclosure.

Furthermore, the computer functions as an AI-CAD section 642 by the processor 602 executing the instructions of the diagnostic support program 640.

The AI-CAD unit 642 uses the master model MM or the local model LM as an inference model and outputs an inference result for input data. The input data to the AI-CAD unit 642 is, for example, a medical image such as a two-dimensional image, a three-dimensional image, or a moving image, and the output from the AI-CAD unit 642 is, for example, information indicating the position of a lesion site within the image, Alternatively, it may be information indicating a class classification such as a disease name, or a combination thereof.

《Explanation of local learning management program 630》
As already explained, the local learning management program 630 is constructed on the client terminal (client 20) existing in the medical institution network 50. The client terminal here may be, for example, the CAD server 60 in FIG. 2. This local learning management program 630 synchronizes the master model MM and local model LM before learning, starts local learning, sets end conditions for local learning, and sends the results of local learning to the integrated server 30 when local learning ends. It has the function of

FIG. 5 is a flowchart illustrating an example of the operation of the client terminal based on the local learning management program 630. The steps of the flowchart shown in FIG. 5 are performed by processor 602 according to instructions from local learning management program 630.

In step S21, the processor 602 of the CAD server 60 synchronizes the local model LM and the master model MM at the time set in the local learning management program 630. Here, the "set time" may be specified as a fixed value, for example, outside of the hospital's testing business hours, or a record of the operating status of the CAD server 60 may be kept and the CAD server 60 is normally used. It may also be set programmatically by determining the amount of time during which it is not being used.

When synchronizing the local model LM and master model MM, for example, the parameter file used by the model may be updated and the program reads it to proceed with learning, or a virtual container image or the like may be updated on the integrated server 30 side. It may also be possible to centrally manage the data and deploy the virtual container image on the terminal side, which is the client 20. Through this synchronization process, the master model MM becomes a learning model (local model LM) in an initial state before the start of learning.

In step S22, the processor 602 executes local learning using the local data LD. The local model LM synchronized with the master model MM has its learning process activated by the local learning management program 630, and local learning proceeds with reference to local data LD within the medical institution network 50.

In step S23, the processor 602 determines whether the learning end condition is satisfied. Here, examples of the learning end conditions include the following conditions.

[Example 1] Specify the number of iterations in advance, and end learning after the specified number of iterations.

[Example 2] Inference accuracy is calculated by retaining verification data in the medical institution network 50 and comparing the accuracy of the inference results obtained by inputting the verification data to a well-trained model with respect to the correct answer. , performs learning until it achieves a specified percentage improvement in accuracy. That is, the inference accuracy of the learning model is calculated using the verification data, and the learning is terminated when a specified percentage of accuracy improvement is achieved.

[Example 3] A time limit is set, learning is performed within the time limit, and learning is terminated when the time limit is reached.

Any one of the above-mentioned [Example 1] to [Example 3] may be defined as the termination condition, or the termination condition may be a logical product (AND) or a logical sum (OR) of a plurality of conditions.

If the determination result in step S23 is No, the processor 602 returns to step S22 and continues the local learning process. On the other hand, if the determination result in step S23 is Yes, the processor 602 proceeds to step S24 and ends the learning.

After the learning is completed, the processor 602 transmits the learning results to the integrated server 30 in step S25. For example, the processor 602 saves the learned model in a file and transmits it to the integrated server 30 via the wide area communication line 70.

Each of the plurality of CAD servers 60 shown in FIG. 2 executes machine learning of each local model LM using data stored in PACS servers 58 in different medical institution networks as learning data. The results are sent to the integrated server 30 via the wide area communication line 70.

《Explanation of learning client combination optimization program 33》
FIG. 6 is a flowchart showing an example of the operation of the integrated server 30 based on the learning client combination optimization program 33. The steps of the flowchart shown in FIG. 6 are executed by the processor 302 according to the instructions of the learning client combination optimization program 33.

In step S31, the processor 302 receives learning results from each client 20.

In step S32, the processor 302 extracts a specified number of clients 20 from the client population, and creates a client cluster that is a combination of the extracted clients 20. One client cluster is understood as one pattern of combinations of clients 20. Then, in step S33, the processor 302 integrates the learning results of the clients 20 belonging to the client cluster to create a master model candidate.

Note that in step S32, the combination of the clients 20 constituting the client cluster may be randomly extracted, but more preferably, if the previous combination search status is saved, subsequent searches are performed from that combination. resume.

Further, in step S33, when creating a master model candidate MMC for a certain client cluster client 20, the processor 302 indicates a correspondence relationship indicating which combination of clients 20 was used to create the master model candidate MMC. The information is stored in a data storage unit such as database 36.

In step S34, the processor 302 evaluates the inference accuracy of the created master model candidate MMC. That is, the processor 302 causes the master model candidate MMC to perform inference using the verification data TD existing in the integrated server 30 as input, calculates the inference accuracy, and calculates the inference accuracy and the accuracy target value. compare. Further, the processor 302 stores the calculated inference accuracy and the comparison result between the inference accuracy and the accuracy target value in the database 36 in association with (linking) the master model candidate.

For the inference accuracy of the master model candidate to be compared with the accuracy target value in the process of step S34, an appropriate value from an instantaneous value or a statistical value such as an average value or a median value is used. An example of the processing content of the inference accuracy evaluation applied to step S34 will be described later using FIG. 7.

In step S35, the processor 302 determines whether a master model candidate MMC exceeding the accuracy target value has been obtained.

If the determination result in step S35 is No, that is, if a master model candidate MMC whose inference accuracy exceeds the accuracy target value has not been obtained, the processor 302 proceeds to step S36.

In step S36, the processor 302 determines whether the number of iterations of the combination search has reached the upper limit number of iterations. If the determination result in step S36 is No, the processor 302 returns to step S32, changes the combination of clients 20, and repeats steps S32 to S36. The processor 302 repeats steps S32 to S36 until the upper limit number of iterations is reached or a combination is found for which the inference accuracy of the master model candidate exceeds the accuracy target value. Note that the order of determination in step S35 and step S36 may be reversed.

On the other hand, if the determination result in step S35 is Yes, that is, if the inference accuracy of the master model candidate exceeds the accuracy target value, the processor 302 ends learning (step S37) and proceeds to step S38.

In step S38, the processor 302 sets the master model candidate whose inference accuracy exceeds the accuracy target value as the latest model with improved performance after learning, and stores this model in a data storage unit such as the database 36 in an appropriate format such as a file. , notifies you that learning has finished. Here, a message queue or general inter-process communication can be used as the notification method. Note that the notification that the learning has ended may be displayed on the display device 316 or may be sent to the client 20.

If the determination result in step S36 is Yes, that is, if no combination of clients 20 for which the inference accuracy of the master model candidate MMC exceeds the accuracy target value is found within the upper limit number of iterations, the processor 302 proceeds to step S39.

In step S39, the processor 302 sets the master model candidate MMC with the best inference accuracy found in the repetition of steps S32 to S36 as a provisional master model, synchronizes this model with the local model LM of the client, and synchronizes this model with the local model LM of the client, as shown in FIG. Steps S21 to S25 and steps S31 to S39 in FIG. 6 are repeated.

《Example of inference accuracy evaluation process》
FIG. 7 is a flowchart illustrating an example of a process for evaluating the inference accuracy of the master model candidate MMC in the integrated server 30. The flowchart shown in FIG. 7 is applied to step S34 in FIG. Although the inference accuracy evaluation process for one master model candidate MMC will be described here, the same process is performed for each master model candidate MMC created from each of a plurality of client clusters with different combinations of clients 20.

In step S341 of FIG. 7, the processor 302 causes the master model candidate MMC to perform inference using the verification data TD as input.

In step S342, the processor 302 calculates the inference accuracy of the master model candidate MMC based on the inference result and the correct data.

In step S343, the processor 302 compares the inference accuracy of the master model candidate MMC with the accuracy target value. Although the accuracy target value may be compared with the instantaneous inference accuracy value of the master model candidate MMC, steps S31 to S343 may be performed while keeping the configuration of the client cluster used for creating the master model candidate MMC fixed. It is also possible to perform the procedure several times, record the inference accuracy each time, and compare statistical values such as the average value and median of the inference accuracy with the accuracy target value.

In step S344, the processor 302 stores the inference accuracy of the master model candidate MMC and the comparison result between the inference accuracy and the accuracy target value in the database 36.

After step S344, the processor 302 ends the flowchart of FIG. 7 and returns to the flowchart of FIG. 6.

<<Specific example of processing by cooperation between the integrated server 30 and multiple clients 20>>
A more specific example of the processing performed by the integrated server 30 and the plurality of clients 20 will now be described. Here, the plurality of clients 20 are assumed to be the plurality of CAD servers 60 shown in FIG. The integrated server 30 and the plurality of CAD servers 60 execute the following processes [Procedure 301] to [Procedure 307].

[Step 301] A client program for distributed learning is executed on the CAD server 60 in the medical institution network 50 of each of a plurality of medical institutions.

[Step 302] The integrated server 30 randomly extracts some client groups (client clusters) to be used for learning from a client group including countless clients 20, which is a client population, and selects different combinations of clients 20. Create multiple client clusters.

[Step 303] The distributed learning clients 20 in each client cluster calculate the number of iterations set for learning using data (for example, medical images) and associated information in the medical institution network to which each client cluster belongs. conduct.

[Step 304] Each client 20 transmits the learned weight parameters of the learning model to the integrated server 30 via the wide area communication line 70.

[Step 305] The integrated server 30 aggregates the weight parameters of the learning results sent from the clients 20 for each client cluster, and creates a master model candidate MMC for each client cluster.

[Step 306] The integrated server 30 verifies the accuracy of each master model candidate created for each client cluster. The integrated server 30 causes the master model candidate MMC to perform inference regarding the verification data TD, and compares the inference result with the correct data.

[Step 307] The integrated server 30 checks the inference accuracy of the model with the highest inference accuracy among the master model candidate MMCs created for each client cluster. If the highest inference accuracy exceeds the target accuracy (accuracy target value), the master model candidate MMC with the highest accuracy (the highest inference accuracy) is adopted as the product model.

On the other hand, if the inference accuracy of the master model candidate MMC with the highest accuracy is lower than the accuracy target value, the integrated server 30 sets the model inference accuracy as an objective function and performs the weight integration of the client 20 that maximizes the model inference accuracy. Search for combinations within a specified time.

For example, the average of the learning results of clients CL1, CL3, and CL5 was used as the weight parameter of the master model candidate MMC, but it was changed to a combination of clients CL1, CL3, and CL6, and the average of these learning results was used as the weight parameter of the master model candidate MMC. Exploring combinations, such as trying different things.

As a result, if a weight combination for the client 20 is found that yields a master model candidate MMC with an inference accuracy exceeding the accuracy target value within the search limit time, that master model candidate MMC is adopted as the product model.

On the other hand, if a weight combination for the client 20 that yields a master model candidate MMC with an inference accuracy exceeding the accuracy target value within the search time limit is not found, the integrated server 30 selects one of the various combinations tried during the search process. The learning iterations from step 303 to step 307 are performed again using the client cluster with higher accuracy.

The integrated server 30 repeats the learning from step 303 to step 307 until a master model candidate MMC with inference accuracy exceeding the accuracy target value is obtained. Alternatively, if a master model candidate MMC with an inference accuracy exceeding the accuracy target value is not obtained even after repeated iterations up to the specified upper limit number of iterations, the integrated server 30 performs the maximum inference in the search process up to that point. The master model candidate MMC for which accuracy has been obtained may be adopted as a product model.

In this way, the new master model created by implementing the machine learning method using the machine learning system 10 according to the present embodiment has improved inference accuracy compared to the master model before learning.

According to this embodiment, it is possible to update the inference performance of the master model MM. When providing a new master model created by implementing the machine learning method according to this embodiment through sales, etc., the number of clients used for learning and the number of clients used for accuracy verification shall be included in the attached document at the time of sale. It is preferable that the number of verification data obtained is described. Regarding the number of clients used for learning, we categorized the clients as a client summary, for example, "how many hospitals," "how many clinics with beds," and "how many clinics without beds." It is preferable to indicate.

In addition, as a preliminary procedure when upgrading from the current product master model, information specifying the inference accuracy of the previous version and the new version, as well as the number and classification of clients used for additional learning. This information is also presented to the medical institution and the medical institution obtains approval before upgrading. After receiving approval, the version will be updated.

《Example of computer hardware configuration》
FIG. 8 is a block diagram showing an example of the hardware configuration of a computer. Computer 800 may be a personal computer, a workstation, or a server computer. The computer 800 can be used as a device having some or all of the previously described client 20, integrated server 30, PACS server 58, CAD server 60, and terminal 62, or a plurality of these functions.

The computer 800 includes a CPU (Central Processing Unit) 802, a RAM (Random Access Memory) 804, a ROM (Read Only Memory) 806, a GPU (Graphics Processing Unit) 808, a storage 810, a communication unit 812, an input device 814, and a display device 816. and a bus 818. Note that the GPU 808 may be provided as necessary.

The CPU 802 reads various programs stored in the ROM 806, storage 810, etc., and executes various processes. RAM804 is used as a work area for CPU802. Further, the RAM 804 is used as a storage unit that temporarily stores read programs and various data.

The storage 810 includes, for example, a hard disk device, an optical disk, a magneto-optical disk, a semiconductor memory, or a storage device configured using an appropriate combination thereof. The storage 810 stores various programs, data, etc. necessary for inference processing and/or learning processing. A program stored in storage 810 is loaded into RAM 804, and CPU 802 executes the program, thereby causing computer 800 to function as a means for performing various processes specified by the program.

The communication unit 812 is an interface that performs communication processing with an external device by wire or wirelessly, and exchanges information with the external device. The communication unit 812 can play the role of an information acquisition unit that receives input such as images.

The input device 814 is an input interface that accepts various operational inputs to the computer 800. Input device 814 may be, for example, a keyboard, a mouse, a touch panel, or other pointing device, a voice input device, or any suitable combination thereof.

The display device 816 is an output interface on which various information is displayed. The display device 816 may be, for example, a liquid crystal display, an organic electro-luminescence (OEL) display, a projector, or an appropriate combination thereof.

《About the programs that run the computer》
At least one of various processing functions such as the local learning function in each client 20 described in the above embodiment, and the learning client combination optimization function including the master model candidate creation function and the inference accuracy evaluation function in the integrated server 30. A program that causes a computer to perform some or all of the processing functions is recorded on a computer-readable medium that is a non-transitory information storage medium such as an optical disk, a magnetic disk, or a semiconductor memory, and the program is executed through this information storage medium. It is possible to provide

Furthermore, instead of providing the program by storing it in a tangible, non-transitory computer-readable medium, it is also possible to provide the program signal as a download service using a telecommunications line such as the Internet.

In addition, part or all of at least one processing function among the plurality of processing functions including the local learning function, learning client combination optimization function, and inference accuracy evaluation function described in each of the above embodiments is provided as an application server. It is also possible to provide services that provide processing functions over telecommunications lines.

《About the hardware configuration of each processing unit》
Master model storage unit 320, verification data storage unit 322, client combination optimization processing unit 330, client cluster extraction unit 332, master model candidate generation unit 334, inference accuracy evaluation unit 340, inference unit 342, inference accuracy shown in FIG. 3 Calculation unit 344, accuracy target value comparison unit 346, display control unit 350, synchronization processing unit 631 shown in FIG. 4, learning data acquisition unit 632, local model LM, error calculation unit 634, optimizer 635, learning result storage unit 636, transmission The hardware structure of processing units (processing units) such as the processing unit 637, the AI-CAD unit 642, and the display control unit 650 that execute various processes includes, for example, various processors as shown below. It is.

Various types of processors include the CPU, which is a general-purpose processor that executes programs and functions as various processing units, the GPU, which is a processor specialized in image processing, and the FPGA (Field Programmable Gate Array), which have circuit configurations after manufacturing. A dedicated electrical circuit that is a processor with a circuit configuration specifically designed to execute a specific process, such as a programmable logic device (PLD), which is a processor that can change the process, and an ASIC (Application Specific Integrated Circuit). etc. are included.

One processing unit may be composed of one of these various types of processors, or may be composed of two or more processors of the same type or different types. For example, one processing unit may be configured by a plurality of FPGAs, a combination of a CPU and an FPGA, or a combination of a CPU and a GPU. Further, the plurality of processing units may be configured with one processor. As an example of configuring multiple processing units with one processor, first, one processor is configured with a combination of one or more CPUs and software, as typified by computers such as clients and servers. There is a form in which a processor functions as multiple processing units. Second, there are processors that use a single IC (Integrated Circuit) chip, as typified by System On Chip (SoC), which implements the functions of an entire system including multiple processing units. be. In this way, various processing units are configured using one or more of the various processors described above as a hardware structure.

Furthermore, the hardware structure of these various processors is more specifically an electric circuit (circuitry) that is a combination of circuit elements such as semiconductor elements.

《Advantages of this embodiment》
The machine learning system 10 according to the embodiment of the present invention has the following advantages.

[1] Learning can be performed without taking personal information such as diagnostic images, which requires privacy considerations, from the medical institution.

[2] An optimal combination can be found from among the multiple clients 20 to quickly improve the inference accuracy of the model. Therefore, even if there is a bias in the learning environment of each client 20, it is possible to achieve the target inference accuracy relatively quickly.

[3] In federated learning, a mechanism is provided for optimizing the combination of clients 20 used when creating a new model by integrating learning results. This makes it possible to quickly achieve high inference accuracy compared to methods that integrate the learning results of all clients 20 or randomly extract combinations from the client population, and allows learning to reach the target accuracy. The time required can be shortened.

[4] It becomes possible to create an AI model with high inference accuracy.

《Modification 1》
In the above embodiment, an AI model for medical image diagnosis was used as an example, but the scope of application of the technology of the present disclosure is not limited to this example. For example, an AI model that uses time series data as input data or an AI model that uses time series data as input data It can also be applied when learning an AI model that uses document data. The time series data may be, for example, electrocardiogram waveform data. The document data may be, for example, a diagnostic report, and can be applied to learning an AI model that supports report creation.

《Modification 2》
In the above-described embodiment, an example has been described in which an accuracy target value by learning is determined and the inference accuracy of a master model candidate is compared with the accuracy target value, but the accuracy target value may be updated as necessary. Alternatively, combinatorial optimization may be performed without predetermining an accuracy target value, under the condition that the inference accuracy of the model is maximized within a time limit or within a specified number of iterations.

"others"
The configurations described in the above-described embodiments and the items described in the modified examples can be used in appropriate combinations, and some items can also be replaced. The present invention is not limited to the embodiments described above, and various modifications can be made without departing from the spirit of the present invention.

10 Machine learning system 20 Client 30 Integrated server 33 Client combination optimization program for learning 36 Database 50 Medical institution network 52 CT device 54 MRI device 56 CR device 58 PACS server 60 CAD server 62 Terminal 64 Local communication line 70 Wide area communication line 80 Learning Data storage unit 302 Processor 304 Computer readable medium 306 Communication interface 307 Procedure 308 Input/output interface 310 Bus 314 Input device 316 Display device 320 Master model storage unit 322 Verification data storage unit 324 Synchronization program 330 Client combination optimization processing unit 332 Client cluster Extraction unit 334 Master model candidate creation unit 340 Inference accuracy evaluation unit 342 Inference unit 344 Inference accuracy calculation unit 346 Accuracy target value comparison unit 350 Display control unit 602 Processor 604 Computer readable medium 606 Communication interface 608 Input/output interface 610 Bus 614 Input device 616 Display device 630 Local learning management program 631 Synchronization processing section 632 Learning data acquisition section 634 Error calculation section 635 Optimizer 636 Learning result storage section 637 Transmission processing section 640 Diagnosis support program 642 AI-CAD section 650 Display control section 800 Computer 802 CPU
804 RAM
806 ROM
808 GPU
810 Storage 812 Communication unit 814 Input device 816 Display device 818 Buses CL1 to CL4, CLN, CLN+1 Client LD Local data LM Local model MM Master model MMC Master model candidate TD Verification data S21 to S25 Steps S31 to S39 of local learning management processing Steps S341 to S344 of learning client combination optimization processing Steps of inference accuracy evaluation processing

Claims (25)

A machine learning system including multiple client terminals and an integrated server,
Each of the plurality of client terminals is
a learning processing unit that executes machine learning of a learning model using data stored in a data storage device of a medical institution as learning data;
a transmitting unit that transmits learning results of the learning model to the integrated server,
The integrated server is
A trained master model and
a synchronization processing unit that synchronizes the learning model on each client terminal side and the master model before causing the plurality of client terminals to learn each of the learning models;
a receiving unit that receives each of the learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; a client combination optimization processing unit that performs at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of model candidates;
Equipped with
The client combination optimization processing unit creates a combination of the client terminals after receiving the learning results from each of the plurality of client terminals,
integrating the learning results of the combination of client terminals to create the master model candidate;
Machine learning system. A machine learning system including multiple client terminals and an integrated server,
Each of the plurality of client terminals includes:
a learning processing unit that executes machine learning of a learning model using data stored in a data storage device of a medical institution as learning data;
a transmitting unit that transmits learning results of the learning model to the integrated server,
The integrated server is
A trained master model and
a synchronization processing unit that synchronizes the learning model on each client terminal side and the master model before causing the plurality of client terminals to learn each of the learning models;
a receiving unit that receives each of the learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; a client combination optimization processing unit that performs at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of model candidates;
Storing information indicating a correspondence between the inference accuracy of the master model candidate created for each combination of the client terminals and the combination of the client terminals used to create the master model candidate. an information storage section to
A machine learning system equipped with The client combination optimization processing unit includes:
a client cluster creation unit that creates a client cluster that is a combination of the client terminals from the plurality of client terminals;
a master model candidate creation unit that creates the master model candidate by integrating the learning results of the client cluster;
an accuracy evaluation unit that detects a master model candidate whose inference accuracy exceeds an accuracy target value by evaluating the inference accuracy of the master model candidate;
The machine learning system according to claim 1 or 2 , comprising: The accuracy evaluation section includes:
an inference accuracy calculation unit that calculates inference accuracy of the master model candidate by inputting verification data into the master model candidate and comparing an inference result output from the master model candidate with correct data of the verification data; and,
an accuracy target value comparison unit that compares the inference accuracy of the master model candidate and the accuracy target value;
The machine learning system according to claim 3 , comprising: The accuracy evaluation section includes:
Based on a comparison between an instantaneous value of inference accuracy of the master model candidate and the accuracy target value, or based on a comparison between a statistical value of inference accuracy in each learning iteration of the master model candidate and the accuracy target value. 5. The machine learning system according to claim 3 , wherein the machine learning system determines whether the inference accuracy of the master model candidate exceeds the accuracy target value. The client combination optimization processing unit includes:
a process of extracting a specified number of the client terminals from the plurality of client terminals to create a combination of the client terminals;
a process of integrating the learning results for each combination of the client terminals to create the master model candidate for each combination;
A process of searching for a combination of the client terminals that exceeds the accuracy target value based on a comparison result between the inference accuracy of each of the master model candidates created for each combination and the accuracy target value;
The machine learning system according to any one of claims 1 to 5 , which performs the following. The machine learning system according to any one of claims 1 to 6 , wherein each of the plurality of client terminals is a terminal installed in a medical institution network of a different medical institution. The machine learning system according to any one of claims 1 to 7 , wherein the integrated server is installed within a medical institution network or outside a medical institution network. The machine learning system according to any one of claims 1 to 8 , wherein the learning result sent from the client terminal to the integrated server includes weight parameters of the learning model after learning. The data used as the learning data includes at least one type of data among two-dimensional images, three-dimensional images, moving images, time series data, and document data.
A machine learning system according to any one of claims 1 to 9 . The machine learning system according to any one of claims 1 to 10 , wherein each of the learning model, the master model, and the master model candidate is configured using a neural network. The data used as the learning data includes a two-dimensional image, a three-dimensional image, or a moving image,
Each model of the learning model, the master model, and the master model candidate is
Machine learning system according to any one of claims 1 to 11 , configured using a convolutional neural network. The data used as the learning data includes time series data or document data,
The machine learning system according to any one of claims 1 to 11 , wherein each of the learning model, the master model, and the master model candidate is configured using a recurrent neural network. The integrated server is
The machine learning system according to any one of claims 1 to 13, further comprising a display device that displays inference accuracy in each learning iteration of each of the master model candidates created for each combination of the client terminals. further comprising a verification data storage unit storing verification data used when evaluating the inference accuracy of the master model candidate;
A machine learning system according to any one of claims 1 to 14. A machine learning method using multiple client terminals and an integrated server,
Before causing each of the plurality of client terminals to learn the learning model, synchronizing the learning model on each client terminal side with a learned master model stored in the integrated server;
Each of the plurality of client terminals executes machine learning of the learning model using data stored in respective data storage devices of different medical institutions as learning data;
each of the plurality of client terminals transmitting learning results of the learning model to the integrated server;
The integrated server,
receiving each of the learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; performing at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of the model candidate;
including;
The integrated server,
creating a combination of the client terminals after receiving each of the learning results from the plurality of client terminals;
integrating the learning results of the combination of client terminals to create the master model candidate;
Machine learning methods. An integrated server connected to multiple client terminals via a communication line,
a master model storage section that stores the trained master model;
a synchronization processing unit that synchronizes the learning model on each client terminal side and the master model before making the plurality of client terminals learn their respective learning models;
a receiving unit that receives learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; a client combination optimization processing unit that performs at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of model candidates;
Equipped with
The client combination optimization processing unit creates a combination of the client terminals after receiving the learning results from each of the plurality of client terminals,
integrating the learning results of the combination of client terminals to create the master model candidate;
Integrated server. An integrated server connected to multiple client terminals via a communication line,
a first processor;
a first computer-readable medium, which is a non-transitory tangible object, on which a first program executed by the first processor is recorded;
The first processor follows instructions of the first program,
storing a trained master model in the first computer-readable medium; and before causing the plurality of client terminals to learn their respective learning models, the learning model on each client terminal side and the master model are to synchronize the
receiving learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; performing at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of the model candidate;
Execute processing including
the first processor creates a combination of the client terminals after receiving each of the learning results from the plurality of client terminals;
executing a process of integrating the learning results of the combination of client terminals to create the master model candidate;
Integrated server. The first processor follows instructions of the first program,
extracting some of the client terminals from the plurality of client terminals and creating a client cluster that is a combination of the client terminals;
creating a master model candidate by integrating the learning results of the client cluster;
The integrated server according to claim 18, wherein the integrated server executes processing including: A machine learning method using multiple client terminals and an integrated server,
Before causing each of the plurality of client terminals to learn the learning model, synchronizing the learning model on each client terminal side with a learned master model stored in the integrated server;
Each of the plurality of client terminals executes machine learning of the learning model using data stored in respective data storage devices of different medical institutions as learning data;
each of the plurality of client terminals transmitting learning results of the learning model to the integrated server;
The integrated server,
receiving each of the learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; performing at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of the model candidate;
Storing information indicating a correspondence between the inference accuracy of the master model candidate created for each combination of the client terminals and the combination of the client terminals used to create the master model candidate. to do and
machine learning methods, including; An integrated server connected to multiple client terminals via a communication line,
a master model storage section that stores the trained master model;
a synchronization processing unit that synchronizes the learning model on each client terminal side and the master model before making the plurality of client terminals learn their respective learning models;
a receiving unit that receives learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; a client combination optimization processing unit that performs at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of model candidates;
Storing information indicating a correspondence between the inference accuracy of the master model candidate created for each combination of the client terminals and the combination of the client terminals used to create the master model candidate. an information storage section to
An integrated server with A program for making a computer function as an integrated server connected to multiple client terminals via a communication line,
to the computer;
A function to save the trained master model,
a function of synchronizing the learning model on each client terminal side and the master model before making the plurality of client terminals learn their respective learning models;
a function of receiving learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; a function of performing at least one of the processes of searching for a combination of the client terminals that maximizes the inference accuracy of the model candidate;
Realize,
a function for the computer to create a combination of the client terminals after receiving the learning results from each of the plurality of client terminals;
a function of integrating the learning results of the combination of client terminals to create the master model candidate;
A program to make this happen . A program for making a computer function as an integrated server connected to multiple client terminals via a communication line,
to the computer;
A function to save the trained master model,
a function of synchronizing the learning model on each client terminal side and the master model before making the plurality of client terminals learn their respective learning models;
a function of receiving learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; a function of performing at least one of the processes of searching for a combination of the client terminals that maximizes the inference accuracy of the model candidate;
Storing information indicating a correspondence between the inference accuracy of the master model candidate created for each combination of the client terminals and the combination of the client terminals used to create the master model candidate. and the ability to
A program to make this happen. A method for creating an inference model by performing machine learning using multiple client terminals and an integrated server, the method comprising:
Before causing each of the plurality of client terminals to learn the learning model, synchronizing the learning model on each client terminal side with a learned master model stored in the integrated server;
Each of the plurality of client terminals executes machine learning of the learning model using data stored in respective data storage devices of different medical institutions as learning data;
each of the plurality of client terminals transmitting learning results of the learning model to the integrated server;
The integrated server,
receiving each of the learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; performing at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of the model candidate;
The master model candidate that achieved the inference accuracy that satisfies the target accuracy, or the client that was used to create the model with the highest inference accuracy among the plurality of master model candidates created by the search process. Creating the inference model having higher inference accuracy than the master model based on the master model candidate created using the combination of terminals;
including;
The integrated server,
creating a combination of the client terminals after receiving each of the learning results from the plurality of client terminals;
integrating the learning results of the combination of client terminals to create the master model candidate;
How to create an inference model. A method for creating an inference model by performing machine learning using multiple client terminals and an integrated server, the method comprising:
Before causing each of the plurality of client terminals to learn the learning model, synchronizing the learning model on each client terminal side with a learned master model stored in the integrated server;
Each of the plurality of client terminals executes machine learning of the learning model using data stored in respective data storage devices of different medical institutions as learning data;
each of the plurality of client terminals transmitting learning results of the learning model to the integrated server;
The integrated server,
receiving each of the learning results from the plurality of client terminals;
a process of searching for a combination of the client terminals in which the inference accuracy of a master model candidate created by integrating the learning results of the combinations of the client terminals that are part of the plurality of client terminals satisfies a target accuracy; performing at least one of the processes of searching for a combination of client terminals that maximizes the inference accuracy of the model candidate;
Storing information indicating a correspondence between the inference accuracy of the master model candidate created for each combination of the client terminals and the combination of the client terminals used to create the master model candidate. to do and
The master model candidate that achieved the inference accuracy that satisfies the target accuracy, or the client that was used to create the model with the highest inference accuracy among the plurality of master model candidates created by the search process. Creating the inference model having higher inference accuracy than the master model based on the master model candidate created using the combination of terminals;
How to create an inference model that includes.

Images